I was unable to join the picture until the post-production phase, when all of the first unit photography had been completed, so I came in with approximately six months left. I was presented with the live action plates as they had been shot and given the job of creating the matte shots to go with those plates. We ended up by having approximately 50 matte paintings in the show.

Initially the problem was that a lot of the equipment we had counted on using had not been completed. We had intended to use front-projection fairly extensively, but the equipment was not finished when I came. In fact, the lamp-house was still being built. Things like that are always a bit worrisome, because you know you are going to have to do a little research and development while you are trying to get the picture done. But we were fortunate; we put in a lot of long hours and had a lot of helpful people and I think we got the job done on a satisfactory level.

My initial worry was that there were a number of plates that had been shot in England and, for one reason or another, weren't as they should be. We had a number of plates that were out of focus; we had a number of plates that had some degree of camera movement in them, a little jiggle. We thought, "Gee, if these plates have such things wrong with them, what other problems might we have that we don't know about yet?" So we just kind of took it a day at a time.

We were on a very tight schedule, which is always the case. You can never have long enough, but that's the nature of the business. We put our faith in the front-projection unit, feeling that it would be the right way to go. As it turned out, we used front projection; we used some rear-projection; we used bi-pack, and we also shot matte paintings onto VistaVision as VistaVision elements to be composited optically on the optical printer. So we used quite a number of methods to get the job done and that was because we found, as we got into each shot, that certain advantages from each method would apply to that shot better than other methods and give us the best possible result. Each method had its disadvantages, certainly, but you can minimize those if you are not worried about that particular disadvantage on that particular shot.

On the several methods that we used (front-projection, rear-projection and bi-pack), we still ended up by using both low-contrast plates and separation masters, depending upon the subject matter. If we wanted to have the contrast slightly increased, as perhaps in the case of a shot that was a little soft, we could pump the contrast and give it a little sharpness. Where we were very concerned about grain, very concerned about reproduction quality, then we would use separations. Generally, that's kind of the way it broke down.

Toward the end of our work on the film, it got a little tense. You don't want to switch horses in midstream too often, because very time you do it sets you back a week or two, while you kind of bring along the new method and do tests and hope to get rid of the gremlins and the glitches. So it becomes a little tricky as the weeks start compressing on you and you begin counting the days. There is a lot of finger-crossing and a lot of hoping that you can stick with a method that is going to pan out. Eventually, if it's not panning out, you've got to make a decision pretty quickly to go to another method and it gets rather interesting. We shipped our last shot on the very last day, literally. We shot it on a weekend and knew that whatever we shot would have to go into the picture. Luckily, it was a good shot. It was a very simple one, as a matter of fact, but you never know which ones are going to be tough.